Deep Learning Prototype Domains for Person Re-Identification

TL;DR

This paper introduces a deep learning approach for person re-identification that automatically discovers prototype domains, enabling scalable and adaptive identification across unseen scenes without requiring domain-specific training data.

Contribution

The work proposes a novel deep learning method for automatic prototype-domain discovery in person re-id, eliminating the need for domain adaptation with target data.

Findings

Outperforms state-of-the-art methods on CUHK-SYSU and PRW benchmarks.

Effective with low-resolution and occluded images.

No need for domain-specific training data.

Abstract

Person re-identification (re-id) is the task of matching multiple occurrences of the same person from different cameras, poses, lighting conditions, and a multitude of other factors which alter the visual appearance. Typically, this is achieved by learning either optimal features or matching metrics which are adapted to specific pairs of camera views dictated by the pairwise labelled training datasets. In this work, we formulate a deep learning based novel approach to automatic prototype-domain discovery for domain perceptive (adaptive) person re-id (rather than camera pair specific learning) for any camera views scalable to new unseen scenes without training data. We learn a separate re-id model for each of the discovered prototype-domains and during model deployment, use the person probe image to select automatically the model of the closest prototype domain. Our approach requires neither supervised nor unsupervised domain adaptation learning, i.e. no data available from the target domains. We evaluate extensively our model under realistic re-id conditions using automatically detected bounding boxes with low-resolution and partial occlusion. We show that our approach outperforms most of the state-of-the-art supervised and unsupervised methods on the latest CUHK-SYSU and PRW benchmarks.